Tuesday, April 19, 2011

Language and Categories

I mentioned in the previous post that we are pattern recognition machines rather than logic machines. Although we can teach ourselves to structure our thoughts logically, it is something of an imposition on our normal way of thinking. Human cognition works mostly like a neural network. We have particular meanings assigned to particular patterns that form something like nodes within our minds, and each node has a collection of other nodes with which it is strongly or weakly connected. The connections can be altered through training ourselves, consciously or not.

For instance, my son enjoys eating apples. He eats them a lot. When I see an apple (which is a node), I now think of my son (another node). Of course, when I think of him, I think of my other son (and another, etc.). And when I think of my other son, I think of the fact that I am taking him to a baseball game (weather permitting) on Sunday, which has nothing at all to do with an apple in any logical sense. However, within my neural network, the apple and the baseball game are only a few nodes apart.

Our minds are set up in an ideal way for keeping ourselves alive in a hostile world. If some ancient man saw tiger tracks or fire, he could relate that to danger. This is not as important for us as it used to be, since we have invented things that do the danger watching for us and we have organized our society in such a way that there are individuals (cops, the army) who theoretically keep us safe in various ways.

In our modern world, this neural network-based way of thinking can sometimes cause problems for us. See, in logic, if we discover that something is wrong, we can simply mark it as such and continue, knowing that. With a neural network, it takes time to unlearn connections that may not be correct. For instance, let's say that as a child I hated going to the doctor, and the doctor's office was painted blue. I have developed a strong association between blue rooms and the misery of the doctor's office. Now in college, I attend a class on a subject in which I am very interested, but it is in a room that is painted blue. Suddenly, for reasons I can't explain and of which I may not be aware, I find it hard to sit in class. When I'm in there, I just want to get out. I stop going to the lectures, and even though the subject is important to me, I do poorly in the class.

This is a rather simplistic example and probably not the best one, but it's the best I can do off the top of my head (the bottom of my head is currently unavailable as it is thinking about chocolate). The point, though, is that in our lives, we develop associations between nodes that are inefficient and potentially harmful to us. We all have a vast collection of such associations, and we're probably only aware of a small number of them.

One aspect of this way of thinking is that it forces us to categorize our experiences. We find patterns and make associations between patterns. We categorize things as good, bad, pleasant, harmful, and so on. Language itself is a means of categorizing the world around us. A particular pattern may be known to us as a tree, while another may be known to us as a chair or a computer or a platypus or a shoggoth. This is not limited to objects. We categorize actions as walking or running or jumping, etc. We understand the world in these terms.

Most people probably consider categorization to be something logical. After all, some of the most famous ones, such as the biological taxonomic ranking system or the Dewey decimal system, appear quite orderly and rational. However, they function more as an attempt to categorize and rationalize systems with no inherent classification or division by minds unable to process continuous data meaningfully. We've all come across books that don't fit the classifications well, and there have been new branches added to various levels of the taxonomic ranking system in order to cope with specimens that fall between the cracks since its invention.

Categorization is an abstraction and simplification of reality. It is an attempt to take something that is continuous and complex - which we have trouble with - and make it into something with clear divisions. That is, something with nodes. We have a tendency to think of the categories we have made to help us understand reality as part of the underlying reality itself, and much error and misunderstanding has resulted from this.

Don't get me wrong. Categorical thinking is important to us. We can't really go without it. We need abstraction and simplification in order to make sense of the world. However, we should not make the mistake of confusing the reality with the categories into which we have shoehorned it. There are no chairs. There are no tables. These are categories we have constructed for certain patterns we encounter. Taken to its logical extreme, one could say that there are no people, either. Indeed, if we try to understand reality continuously rather than categorically, it seems to go against some of our most strongly held ideas and beliefs.

But conclusions reached via categorical thinking are not inherently incorrect. Not at all. Like I said, we are able to detect patterns very well due to our thinking style. We can detect a tendency for the trees to sway when we hear wind. We can detect a tendency for tones to sound when we vibrate a string. These have fed our inventiveness and allowed us to accomplish all sorts of things. We simply have to understand that we created these categories - they do not actually exist.

No comments:

Post a Comment